2003 Project Summaries
Recipients of Three-year Awards
Xandra O. Breakefield, Ph.D. Effective gene therapy for brain tumors will require an expanded range of gene delivery and selective targeting of invasive tumor cells. This research will focus on these needs at three levels. First, novel forms of the tumor apoptotic TRAIL protein are designed that may be secreted from cells in a controlled manner. This will be achieved by fusing the extracellular domain of TRAIL with protein elements that promote secretion and an ER retention signal joined by a protease recognition site. Thus, TRAIL will be retained in a non-toxic intracellular state until it is activated by proteolytic cleavage. This delay in activation will maximize the use of a second component, neuroprecursor cells (NPCs), which are resistant to TRAIL, home to tumor foci in the brain, and deliver this therapeutic protein. NPCs will be transduced ex vivo and tumor cells in vivo using HSV amplicon vectors that express various forms of TRAIL and viral proteases, ultimately with all components in the same vector and with protease activity under a drug-regulated promoter. Third, effectiveness of gene delivery and tumor regression will be visualized over time in living animals by dual bioluminescence detection of glioma cells that express Renilla luciferase (Rluc) to monitor tumor volume, and vectorsor NPCs that express firefly luciferase (Fluc) to monitor transgene expression and cell migration. This research combines novel therapeutic proteins, modes of gene delivery, and imaging modalities to optimize gene therapy for experimental brain tumors in a manner that is compatible with eventual clinical trials.
Arnab Chakravarti, M.D. Over the decades, radiation (RT) resistance of malignant gliomas has posed a serious clinical problem that lacks an effective solution. Simple radiation dose escalation is complicated by proximity to critical normal brain structures that have limited tolerance to RT. The underlying molecular mechanisms protect normal brain cells vs. glioma cells from RT-induced damage. Resolving these issues is a critical first step toward the development of novel biotherapeutic strategies that can effectively enhance the therapeutic gain of RT. Our preliminary data suggests that the phosphoinositol 3-kinase (PI3K) pathway may play a critical role in protecting normal brain cells and glioma cells from RT-induced cellular damage. Inhibiting PI3K itself would appear to be inadvisable, as increased RT-induced damage would be observed in both gliomas and normal brain cells by very different mechanisms and mediators. We have already identified some of these differential PI3K mediators. Our hypothesis is that we can substantially enhance the therapeutic gain of RT by selectively inactivating PI3K signaling in malignant gliomas (vs. normal brain cells). We have three specific aims. 1) Optimize conventional radiotherapeutic delivery to minimize PI3K activation in gliomas while maximizing PI3K activation in normal brain cells. 2) Develop biotherapeutic approaches to simultaneously and selectively suppress PI3K activation in malignant gliomas (TGF?, MnSOD, cis-ACBS, etc.) to improve the therapeutic ratio of RT. 3) Investigate the role of PI3K pathway members as potential biomarkers in association with biomarkers functioning in other pathways. This last aim will help us a) stratify the patients who are most likely to benefit from convention RT in order to quickly direct those with radioresistent tumors to alternative therapies, and b) identify which subgroups of malignant gliomas are most likely to be responsive to available biotherapeutic agents (Iressa, CCI-779, and FTI) presently in clinical trials. We expect that through such molecular profiling efforts, selection of the most appropriate biotherapeutic agents(s) may be administered and tailored to the individual features of a patient's tumor, rather than toward a broad histopathological group, as is currently the standard of care.
Daniel G. Jay, Ph.D. Advances in brain tumor treatment have been limited (Maher et al., 2001). The mean survival time of patients with glioblastoma multiforme (GBM), the most severe subtype of malignant glioma (Berens and Giese, 1999), is 9 to 12 months (Maher et al., 2001). This is due to the failure of all existing therapies to alter the progression of GBM (Louis et al., 2002). The NCI/NINDS Brain Tumor Progress Review Group (BT-PRG) identified the dispersal of GBM tumors as the root source of therapeutic failure (Louis and Posner, 2000) and proposed that future research efforts should focus on proteomic approaches to address dispersal. While several proteins have already been implicated in GBM dispersal (Berens and Giese, 1999; Bolteus et al., 2001), it is likely that there are many other unidentified proteins that play a role in this process. To address these needs, we propose to conduct an unbiased proteomic screen to identify proteins that mediate GBM dispersal. The screen combines flourophore-assisted light activation (FALI), a protein inactivation strategy that targets light energy via fluorescein-labeled antibodies to damage specific proteins in the cancer proteome, with a high-throughput dispersal assay. Our initial screen has implicated the poliovirus receptor (PVR) in tumor cell dispersal, but we will test the hypothesis that PVR is required for GBM dispersal. We will also test the hypothesis that our screen can identify other proteins that are required for GBM dispersal and further validate the proteins that we find in RNA interference and other complementary approaches. An additional strategy is to develop and test neutralizing antibodies that may provide a direct route to therapeutics. The proposed work will provide new avenues to identify and validate targets for drug development to enhance GBM patient prognosis. We have three specific aims: 1) confirm that PVR is required for GBM dispersal, 2) screen for new proteins required for GBM dispersal, and 3) validate the role of newly-identified proteins in GBM dispersal. Our proposal combines several innovative technologies to address a key priority in GBM research: finding new treatments to inhibit GBM dispersal. To our knowledge, our proposed work is the first functional screen for GBM. Our screen for sarcoma invasion has led to the validation of three new targets and the filing of patent applications. This bodes well for our hopes to identify and validate new proteins that are required for GBM dispersal. The first of these is PVR, and the complementary validation and development of neutralizing scFvs will provide a direct route to therapeutics. We are further excited by the possibility that we may identify other proteins that may also provide druggable targets for GBM treatment. Our long-standing efforts at the intersection of neuroscience and cancer and our recent interest in translational work from lab to clinic places us in an excellent position to contribute to GBM research. Funding from The Goldhirsh Foundation will give us the opportunity to grow in this important field. Recipients of One-year Awards
Bruce R. Ksander, Ph.D. The overall goal of this project is to develop a novel immunotherapy that may ultimately be translated into a treatment for adult astrocytic tumors. While immunotherapies for astrocytomas have been actively pursued, the immune-privileged environment of the brain prevents the success of most. We proposed the use of bioactive microvesicles that express membrane Fas ligand (FasL) to accomplish three goals: 1) terminate immune privilege, 2) activate innate and adaptive immunity, and 3) ultimately eliminate brain tumors and protect patients from secondary metastases. Based on previous research in the eye, we believe that a major role for FasL in immune-privileged sites is the regulation of innate immunity; membrane FasL initiates innate immunity, while soluble FasL prevents innate immunity. Our novel immunotherapy will use the two different forms of FasL to regulate the onset and resolution of innate-mediated inflammation. The bioactive microvesicles that express membrane-only FasL (cannot be cleaved) will be used to "turn on" anti-tumor inflammation in the brain, while soluble FasL will be used to "turn off" inflammation and minimize the non-specific tissue damage inherent with inflammation. We believe our approach may be used to develop a potent immunotherapy that is capable of overcoming the local immunosuppressive environment of the brain. Moreover, we believe this therapy may be translated into a successful treatment for astrocytomas that are largely incurable with present therapies.
Andreas C. Kurtz, Ph.D. The long-term goal of this project is to classify astrocytic tumors by their expression of secreted proteins, use this information to establish a comprehensive serum profile of these proteins, and develop convenient tools to diagnose astrocytoma using body-fluid analysis. We hypothesize that astrocytoma can be classified by their expression profile of secreted proteins. This hypothesis is based on the fact that there are more than 4,000 known potentially secreted proteins, of which a distinct subset is expressed at different stages of astrocytoma progression. Some of these secreted proteins may be detected in body fluids such as blood and cerebrospinal liquor. The non-invasive accessibility of proteins secreted by astrocytomas is a valuable diagnostic advantage. One can determine protein expression profiles from a serum sample and then diagnose and classify the tumor. During the pilot study we will develop a database of the secreted proteins found in astrocytomas. The database will contain published information on gene expression profiles of astrocytomas and histopathological and genetic tumor data. We will also validate the in silico selected marker proteins using tissue microarrays and testing serum samples from our brain tumor serum bank. Finally, we will identify profiles of secreted proteins in the serum of astrocytoma patients-independent of expression-profile based predictions. Such protein profiles will be experimentally derived by proteome analysis of serum from astrocytoma patients and controls. This will be the first systematic study of the serum proteome and its changes during astrocytoma growth and progression. Over time, therapeutic responses may be monitored by measuring the serum levels of proteins secreted from astrocytomas. Analyzing a comprehensive matrix of serum proteins may serve as a window on molecular processes that are associated with astrocytic tumors. This insight may provide information about potential molecular targets for individual tumors and their recurring offspring. A matrix of secreted proteins found in the serum or liquor of astrocytoma patients is a valuable tool that can help us diagnose these tumors on a molecular level and eventually lead us to therapeutic success.
Elizabeth A. Maher, M.D., Ph.D. Low-grade astrocytomas are slow-growing malignant brain tumors that morph into glioblastoma multiforme (GBM) in 70% of patients within 5 years of the initial diagnosis. These tumors affect predominantly young patients and, once they morph, behave in a way that is clinically similar to primary GBM: They become essentially resistant to all types of chemotherapy, progress rapidly, and often cause death in less than a year. This transition to GBM is heralded by dramatic increases in the rate of tumor growth and the size of the invasion, induction of angiogenesis, and often wide-spread necrosis. The molecular mechanisms that drive this transition are largely unknown. To understand the molecular underpinnings of this critical transition from low-grade tumor to aggressive GBM-and design therapies to prevent it from happening-we need to generate a comprehensive profile of the genes that mutate in this critical transition. During the proposed study, DNA extracted from paired specimens of low-grade astrocytoma and secondary GBM from individual patients will be subjected to high-resolution gene specific array CGH that employs cDNAs rather than BACs. This approach-combining genomics with bioinformatics technology-is truly revolutionary, enabling scientists to identify with greater precision (and higher resolution) the genetic changes that occur within a cancer cell. This breakthrough platform, which can detect single copy changes at the gene level in a high-throughput manner, coupled with superior clinical samples will help us to identify critical genetic targets for the development of effective therapies that can prevent the deadly transition from low-grade astrocytoma to GBM.
Vladimir P. Torchilin, Ph.D., D.Sc. In earlier studies, we demonstrated that certain non-pathogenic anti-nuclear auto-antibodies (ANA) bind to the surface of a variety of tumor cells, including some astrocytic tumor cells, but not to normal cells. These antibodies, including monANA 2C5, have a unique nucleosome-restricted specificity and recognize tumor cells through their tumor cell surface-bound nucleosomes. This recognition is associated with an anti-tumor activity of some ANAs, since in preliminary experiments monANA 2C5 effectively suppressed the growth of several unrelated murine and human tumors in mice. In contrast to existing anti-tumor monoclonal antibodies that come from immunized mice and are tumor-type specific, monANA 2C5 and similar antibodies appear to be natural antibodies with broad specificity against tumor cells of different types and origins. Such antibodies may also be used as specific ligands for the delivery of various pharmaceutical agents including drugs, drug carriers, and diagnostic moieties to various tumors-including astrocytic tumors. In this proposal we plan to 1) investigate the interaction of monANA 2C5 with a variety of human astrocytic cell lines and identify the lines with the strongest surface binding of monANA 2C5; 2) prepare antibody-bearing fluorescence- or radio-labled liposomes and micelles and investigate their in vitro interactions with select astrocytic cells; and 3) grow the selected tumors (using a cranial window chamber model) in nude mice and investigate the biodistribution, tumor accumulation, and intratumoral localization of fluorescence- or radio-labeled monANA 2C5 and 2C5-bearing liposomes and micelles. The distant goal of this approach is to develop antibody-based targeted compositions that provide the most efficient localization and treatment of astrocytic tumors. |
 |
 |
